COST-Multi-Wall Model (MW)

The multi-wall model gives the path loss as the free space loss added with losses introduced by the walls and floors penetrated by the direct path between transmitter and receiver.

It has been observed that the total floor loss is a function of the number of the penetrated floors. This characteristic is taken into account by introducing an additional empirical correction factor.

The individual penetration losses for the walls (depending on their material parameters) are considered for the prediction of the path loss. Therefore, the multi-wall model can be expressed as follows:

(1)

l_{M W} = l_{F S} + l_{c} + k_{f} l_{f} + \sum_{i = 1}^{N} k_{w i} l_{w i}

where

$l_{F S}$ is the free space loss between transmitter and receiver
$l_{c}$ is the constant loss
$k_{w i}$ is the number of penetrated walls of type i
$k_{f}$ is the number of penetrated floors
$l_{w i}$ is the loss of wall type i
$l_{f}$ is the loss between adjacent floors
$N$ is the number of different wall types

The constant loss in the equation above is a term which results when wall losses are determined from measurement results by using multiple linear regression. Normally it is close to zero. The third summand in the equation represents the total wall loss as a sum of the walls between transmitter and receiver. For practical reasons in ProMan the individual wall loss of the intersected walls is considered.

It is important to notice that the loss factors in the formula express not physical wall losses but model coefficients which are optimized with the measured path loss data. Consequently, the loss factors implicitly include the effect of furniture. However, wave guiding effects are not considered with this model, thus the accuracy is moderate. On the other hand, this model has a low dependency on the database accuracy and because of the simple approach a very short computation time. Therefore, no preprocessing of the building data is needed for the computation of the prediction and no settings have to be adapted for this prediction model.